HAL Id: hal-00890968https://hal.archives-ouvertes.fr/hal-00890968

Submitted on 1 Jan 1992

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Studies on pollen sources for Apis cerana Fabr and Apismellifera L bees at Muzaffarpur, Bihar, India

Mc Suryanarayana, G Mohana Rao, Tsms Singh

To cite this version:Mc Suryanarayana, G Mohana Rao, Tsms Singh. Studies on pollen sources for Apis cerana Fabrand Apis mellifera L bees at Muzaffarpur, Bihar, India. Apidologie, Springer Verlag, 1992, 23 (1),pp.33-46. <hal-00890968>

Original article

Studies on pollen sources for Apis cerana Fabrand Apis mellifera L bees at Muzaffarpur, Bihar, India

MC Suryanarayana* G Mohana Rao TSMS Singh

Khadi and Village Industries Commission, Central Bee Research and Training Institute,1153 Ganeshkhind Road, Pune 411016, India

(Received 6 May 1991; accepted 23 August 1991)

Summary &mdash; At Muzaffarpur, Bihar, in northern India, a total of 40 and 48 plant species served aspollen sources to Apis cerana Fabr and A mellifera L respectively. The major, medium and minorpollen sources were identified. The peak in pollen availability was observed from January to Mayand again in November. The dearth period was July-August in this area. The main differences inpollen foraging between 2 honey bee species were noted. Each of the species had 13 anemophilousplant species as sources of pollen.

Apis mellifera / Apis cerana / pollen plant / foraging / India

INTRODUCTION

Microscopic analysis of pollen loads col-lected by foragers of a honey bee colonyhas been an established melittopalyno-logical method to determine the sourcesof pollen for bee colonies. As early as1908, Betts (1935) initiated investigationsof pollen loads from Apis mellifera L beecolonies in England, to determine the pol-len plants and to identify factors affectingbee constancy towards flowers. Particular-ly in Europe, this aspect was subsequent-ly examined in several studies, amongwhich mention may be made of those bySynge (1947) and Percival (1947) on pol-len loads collected from A mellifera bee

colonies in the UK; Louveaux (1954,1958) from France; Maurizio (1949) fromSwitzerland; Knee and Moeller (1967),and O’Neal and Waller (1984) from theUSA. A classic example of the study ofpollen loads is that of Hodges (1955), whopresented a comprehensive and practicalguide to identifying pollen sources in theUK.

Melittopalynological studies on thesources of pollen in Asia have been veryfew. Systematic investigations have beenmade of the pollen sources on the Mahab-aleshwar plateau in Maharashtra in west-ern India by Deodikar et al (Deodikar,1965). Routine laboratory studies usingmelittopalynological methods have been

*

Correspondence and reprints

made at the Central Bee Research and

Training Institute, Pune, India, to evaluatesources of pollen and nectar for A ceranaFabr bees in different parts of India (Chan-dran and Suryanarayana, 1970; Suryana-rayana, 1975, 1978; Chaudhari, 1977;Singh et al, 1987; Suryanarayana andSingh, 1989). Sharma (1970 a,b) analysedpollen loads collected from Kangra, Hima-chal Pradesh; and Chaturvedi (1973,1977) investigated the pollen sources at

Banthra (Lucknow), Uttar Pradesh.

Muzaffarpur (26°07’ N, 85°24’ E) is an

important area for honey production in Bi-

har, northern India. Many beekeepers fromdistant areas migrate their bee colonies toMuzaffarpur for the honey harvest. In spiteof its importance for beekeeping, very littleinformation is available on bee foragingsources in this area. The Central Bee Re-search and Training Institute, Khadi and

Village Industries Commission, Pune, es-tablished a Field Observation Station at

Muzaffarpur during 1976 for studies on

various aspects of beekeeping. The resultsof melittopalynological investigations on

the sources of pollen are presented here.

MATERIALS AND METHODS

About 40 colonies of A cerana and 30 coloniesof A mellifera have been maintained at the FieldObservation Station, Muzaffarpur. The 2 beespecies were kept in 2 separate apiaries 2 kmapart during 1987-1988, while during 1988-

1989, they were kept in the same apiary about 5km away from the 2 previous apiaries (fig 1). Bo-tanical surveys of the areas around these apiar-ies have been made and observations weremade on the component flowering species, aswell as visits of bees to these flowers. Samplesof pollen (anthers, flowers and inflorescences)of the flowering species were collected. Refer-

ence slides of pollen were prepared followingthe methods of Wodehouse (1935) and Erdt-man (1952).

Samples of pollen loads brought by foragerbees into the colonies were collected through-out the day at hourly intervals. The incomingpollen foragers were caught by hand by holdingtheir wings. The 2 loads on the 2 hind legs weregently removed onto a slip of clean paper; theforager was then let off to resume its activity.The 2 loads were then packed in 2 separatepaper packets. Each packet was labelled withdata on the colony number, bee species, timeand date of collection, colour of the load andnumber of incoming forager bees per minutewith similar type of pollen. This was done at 1-2-hour intervals throughout the day. Samplingwas performed once a week. A total of 810 pol-len loads from A cerana colonies and 1 260

from A mellifera colonies was collected overmore than 2 years of study from January 1987to March 1989. To identify the pollen sources,small amounts of pollen from different parts ofeach pollen load sample were transferred on toa glass slide by a needle and a temporarymount was made. The mount was examined mi-

croscopically for pollen contents. Identificationof the pollen species was confirmed by compar-ing these with the pollen in the reference pollenslides.

RESULTS AND DISCUSSION

Muzaffarpur is a predominantly agriculturalarea. The important annual crops during1987-1989 were Zea mays L, Hordeum

vulgare L, Brassica spp, Helianthus an-

nuus L, Sesamum indicum L, Coriandrumsativum L, Trigonella corniculata L, Caja-nus cajan (L) Millsp, Pisum sativum L vararvense (L) Poir, Vigna sp and differentcucurbit species. These contribute to theseasonal bee forage. Nephelium litchiCamb and Mangifera indica L were impor-tant fruit crops. Other horticultural crops ofminor importance were Psidium guajava L,Citrus spp, Spondias pinnata (Lf) Kurz andEmblica officinalis Gaertn. All these plantsprovided both nectar and pollen, excepting

Emblica officinalis, which is exclusively asource of pollen.

Other pollen and nectar sources were:a), the following arboreal species cultivat-ed or growing wild: Borassus flabellifer L,Bombax ceiba L, Cocos nucifera L, Dal-

bergia sissoo Roxb, Delonix regia (Boj)Raf, Moringa oleifera Lamk, Tamarindus

indica L, Wendlandia exserta (Roxb) DC;b), the most important weeds: Cannabissativa L, Cleome sp, Cyanotis sp, Cosmosbipinnatus Cav, Cyperus sp, and Partheni-um hysterophorus L.

From the results it appeared that a totalnumber of 40 species served as pollensources to A cerana and 48 to A mellifera

(table I). The major pollen sources for

A cerana were, in order of importance,Zea mays, Brassica spp, Phoenix sylves-tris (L) Roxb and Borassus flabellifer. Zeamays provided 34% of the total pollenloads and was a dominant source provid-ing pollen almost every month. Cyanotissp, Cosmos bipinnatus, Fabaceae, Cucur-bitaceae and Cocos nucifera were sources

of medium importance, providing a signifi-cant quantity of pollen.

The major pollen sources for A melliferawere, in order of importance, Zea mays,Parthenium hysterophorus, Brassica spp,Phoenix sylvestris, Borassus flabellifer,Fabaceae, Cucurbitaceae, Cajanus cajan,Pisum sativum var arvense and Cosmos

bipinnatus. However, only Zea mays, Fab-aceae, Cucurbitaceae and Parthenium

hysterophorus provided forage for a majorpart of the year and the others served asforage sources only for a few months. Oth-er sources which contributed significantlyto the total pollen availability, in order of

importance, were Coriandrum sativum, Co-cos nucifera, Cyanotis sp, Helianthus an-nuus and Asteraceae.

In table II a floral calendar is provided ofplants of significance to beekeeping in Mu-zaffarpur. Although pollen is available

year-round in the area studied, the peakpollen availability was during January-Mayand in November. During July, August andSeptember it was poor. A maximum num-ber of sources was visited by A cerana inMay and in November, while in July andAugust the number was the lowest. ForA mellifera the maximum number was in

March, May and November and the lowestwas in July and August. Photomicrographs

of some pollen sources are given in fig-ure 2.

An analysis of the contribution of ane-mophilous plant species to pollen availabil-ity is given in table III. Out of the 40 sourc-

es of pollen for A cerana bees, 13 sourcescan be considered as anemophilous, andcontributed 495 (61.7%) pollen loads. Ofthese 381 (70.7%) loads were contributedby 3 major sources, viz, Borassus flabelli-

fer, Phoenix sylvestris and Zea mays. In

the case of A mellifera also, 13 anemophi-lous species, out of the total number of 48pollen sources contributed 569 (45.2%)pollen loads. Of these, 4 major sources ofpollen, viz, Borassus flabellifer, Phoenix

sylvestris, Parthenium hysterophorus andZea mays, contributed 481 (84.5%) loads.For both honey bee species, anemophi-lous plant species constituted important

pollen sources at Muzaffarpur. Sharma

(1970a) reported a similar predominanceof anemophilous sources of pollen in Kan-gra, Himachal Pradesh, while in the Shim-la hills of the same state, Sharma (1983)recorded only one (Artemisia sp) importantpollen source, which was anemophilous.Zea mays was recorded as the most im-

portant source in southern Wisconsin,USA, by Knee and Moeller (1967), while

wind-pollinated species contributed to anunexpectedly large amount of pollen in theLaCrosse county, Wisconsin, USA (Sever-son and Parry, 1981). Todd and Bretherick(1942), while giving the chemical composi-tion of pollen in 32 plant species, observedthat pollen from anernophilous plants wasnot greatly different from pollens of ento-mophilous plants in their protein value.Maurizio (1960) classified corn pollen un-der the group of highly nutritive pollens.Dietz (1975) mentioned that the differencebetween insect-pollinated and wind-

pollinated plants in protein content of pol-len was non-significant. It is generally ob-served that brood rearing activity in thecolonies is vigorous during the periodwhen maize pollen is available (personalobservations). It therefore appears that an-

emophily as a concept of pollination ecolo-gy is not relevant in the case of social in-sects like honey bees, whose foragingbehaviour is controlled by a different set offactors.

The period and abundance of the mainpollen sources of A cerana and A mellifera

have been shown in figures 3 and 4 re-

spectively.Important differences were found in the

pollen sources between A cerana andA mellifera. These are given below: a) pol-len load samples of A cerana did not haveeven a single load of Parthenium hyster-ophorus. This was noted as the secondmost important pollen source for A mellife-ra. Suryanarayana and Singh (1989) re-

corded earlier that A cerana did not forageon P hysterophorus, although this was acommon weed in the area. b) Samples ofA mellifera pollen loads did not representDendrophthoe falcata (Lf) Ettingsh and

Moringa oleifera, which were sources forA cerana. This might be due to competitionamong the floral sources and their relativeunattractiveness to A mellifera. c) Cajanuscajan and Pisum sativum var arvense wereimportant forage sources for A mellifera,whereas these were only of minor impor-tance for A cerana. This difference in for-

aging behaviour is significant from the

point of view of pollination of these crops.d) Helianthus annuus was an important

source for A meffifera, but it was not repre-sented in the loads brought by A cerana.This might be due to non-availability ofsunflower within the flight range of A cera-na, which has been reported as 700 to 900m (Naim and Phadke, 1972; Darade et al,1989). Suryanarayana et al (1987) showedthat A cerana was one of the importantpollinators of sunflower in Maharashtra.

It is interesting to note that Zea mayswas an important pollen source for bothA cerana and A mellifera in agricultural ar-eas such as Muzaffarpur, as has been ob-served in several parts of the world (for ex-ample, Sladen (1911) in England; Casteel(1912), Knee and Moeller (1967), and Se-verson and Parry (1981) in the USA andMaurizio (1953) in Switzerland). In Indiamaize has been found to be a source of

pollen for A cerana bees in the Punjab

(Chaudhari, 1977) and in Andhra Pradesh(Singh et al, 1987).

Though Nephelium litchi is an importantnectar source (Phadke and Naim, 1974;Suryanarayana et al, 1981; Rao, 1983), itwas found to be a poor pollen source.

Likewise, Mangifera indica and Wendlan-dia exserta, which were reported to benectar sources (Rao and Nair, 1985; andpersonal observations), were not pollensources.

Out of the 810 pollen loads of A ceranaexamined, only one pollen load had a bi-floral source and all the rest were unifloral.The bifloral load was collected in Novem-ber from Phoenix sylvestris and Zea mays.A mellifera also collected one bifloral pol-len load in November from Cucurbitaceaeand Cocos nucifera. Bifloral loads werelikewise found to be rare in the pollen

loads collected by A cerana in Vijayarai,Andhra Pradesh (Singh et al, 1987). How-ever, Sharma (1970a,b) found that 8 pol-len loads out of 19 collected during 1stJune to 9th July, 1968 at Kangra, Hima-chal Pradesh, were mixed; one had 4 pol-len species, 2 had 3 pollen species and 5were bifloral, while only one out of 27 col-lected during September, 1968 was mixedwith 3 pollen species. She therefore con-cluded that June-July was the dearth peri-od in Kangra and that bees were thereforeforced to collect pollen from more than onesource on several of their foraging trips.On the other hand, at Banthra (Lucknow),Uttar Pradesh, she found that 3 out of 49loads collected in February, 1970 and 3out of 44 loads collected in April, 1970were bifloral (Chaturvedi, 1973).

The almost exclusive occurrence of uni-floral loads in the present investigations isindicative of the richness of the local pol-len resources to the honey bees.

ACKNOWLEDGMENTS

Thanks are due to the field staff of the Beekeep-ing Observation Station, Muzaffarpur for theiractive help in the collection of pollen loads andplant specimens during the period of observa-tion. The authors thank the referees of the

paper for their comments and valuable sugges-tions.

Résumé &mdash; Étude sur les sources depollen pour Apis cerana Fabr et Apismellifera L à Muzaffarpur, Bihar (Inde).Muzaffarpur, dans le Bihar (Inde du Nord),est une région principalement agricole.Les cultures constituent les principalessources de pollen et de nectar pour lesabeilles. Au laboratoire de terrain du Cen-tral Bee Research and Training Institute àMuzaffarpur, les pelotes de pollen rappor-tées par les butineuses en 1987-1988 ontété étudiées afin de déterminer l’origine

des pollens. Les pelotes ont été échan-tillonnées toutes les heures durant une

journée entière, une fois par semaine. Ona récolté environ 810 pelotes de pollenprovenant de 40 colonies d’A cerana et1 260 venant de 30 colonies d’A meilifera.En 1987-1988, les colonies des 2 espècesétaient réparties en 2 ruchers distants de2 km, en 1988-1989 elles étaient regrou-pées en un même rucher (fig 1).

Au total, 40 espèces de plantes ont

fourni du pollen à A meilifera et 48 à A ce-rana. Les sources polliniques les plus im-portantes pour A cerana sont : Zea maysL, Brassica spp, Phoenix sylvestris (L)Roxb et Borassus fiabellifer L. À celles-ci il

faut ajouter, pour A meilifera, les espècessuivantes : Parthenium hysterophorus L,Cajanus cajan (L) Millsp, Pisum safivum Lvar arvense (L) Poir et les Cucurbitacées(tableau I). Dans la région de Muzaffarpur,du pollen est disponible toute l’année.C’est de janvier à mai et en novembre qu’ilest le plus abondant; de juillet à septem-bre, la disponibilité est faible (tableau II).Treize plantes anémophiles (61,7% despelotes d’A cerana et 45,2% des pelotesd’A mellifera) constituent des sources depollen importantes pour les 2 espècesd’abeilles (tableau III). Parthenium hystero-phorus représente la 2e source de pollenpour A mellifera, tandis qu’A cerana n’en apas récolté une seule pelote. Les grains depollen de Dendrophthoe falcata (Lf) Et-

tingsh et de Moringa oleifera Lamk, pré-sents dans les pelotes d’A cerana, n’ont

pas été trouvés dans celles d’A mellifera.

Nephelium litchi Camb, principale sourcede nectar à Muzaffarpur, fournit peu de

pollen aux abeilles. Il n’a été trouvé qu’uneseule pelote mixte (2 espèces) dans leséchantillons des 2 espèces d’abeilles. Ceciprouve la richesse des plantes locales

quant à la fourniture de pollen pour ces in-sectes.

Apis mellifera / Apis cerana / plante pol-linifère / Inde / butinage

Zusammenfassung &mdash; Studien über diePollenquellen von Apis cerana Fabrund Apis mellifera L in Muzaffarpur,Bihar (Indien). Muzaffarpur, Bihar, in

Nordindien ist überwiegend ein landwirt-schaftlich genutztes Gebiet. Kultivierte

Nutzpflanzen sind die Hautpollenpflanzenund Nektarpflanzen für die Bienen. An denVölkern von Apis cerana und A melliferader Feldstation des Zentralen Bienenfor-

schungsinstitutes in Muzaffarpur wurdenan den Pollenladungen der heimkehren-den Bienen 1987-1988 Untersuchungenüber die Herkunft des Pollens vorgenom-men. Ein Mal wöchentlich wurden den

ganzen Tag über in stündlichen Abstän-den Pollenproben gesammelt. Etwa von40 Völkern von A cerana und von 30 Völ-kern von A mellifera wurden insgesamt810 Pollenhöschen von ersterer und 1260

Pollenhöschen von letzterer entnommen.

1987-1988 wurden die beiden Bienenar-ten auf zwei getrennten Bienenständen 2km voneinander entfernt gehalten, 1988-1989 aber gemeinsam auf einem Bienen-stand (Abb 1). Insgesamt 40 Pflanzenar-ten lieferten Pollen für A cerana und 48Arten für A mellifera. Die wichtigsten Pol-lenpflanzen für A cerana waren Zea maisL, Brassica spp, Phoenix silvestris L undBorassus flabellifer L. Zusätzlich zu diesendienten für Apis mellifera noch Partheniumhysterophorus L, Cajanus cajan L, Pisumsativum L var arvense L Poir und Cucurbi-taceen als wichtige Pollenquellen (Tab I).In Muzaffarpur ist für die Bienen Pollendas ganze Jahr über vorhanden. Das

Hauptangebot an Pollen lag zwischen

Januar und Mai und im November. VonJuli bis September ist das Pollenangebotknapp (Tab II). Dreizehn windblütige Pflan-zenarten (61.7% der Pollenladungen vonA cerana und 45.2% von A mellifera) bilde-ten für beide Arten wichtige Pollenspender(Tab III). Parthenium hysterophorus warder zweitwichtigste Pollenspender für A

mellifera; aber A cerana sammelte von

dieser Art nicht einmal eine einzigeLadung. Die Pollenkörner von Dendroph-thoe falcata (Lf) Ettingsh und von Moringaoleifera Lamk, die in den Proben von Acerana-Völkern vorhanden waren, wurden

in den Proben von A mellifera-Völkernnicht gefunden. Nephelium litchi Camb,die Hauptnektarquelle in Muzaffarpur,erwies sich als wenig ergiebige Pollen-

quelle. Es wurde bei jeder der beiden Bie-nenarten nur je ein einziges gemischtesPollenhöschen (von zwei Pflanzenarten)gefunden. Das beweist die Reichhaltigkeitder lokalen Pollenpflanzen für die Versor-gung der Bienenvölker.

Apis mellifera / Apis cerana / Pollen-

pflanze / Indien / Trachtverhalten

REFERENCES

Betts AD (1935) The constancy of the pollen-collecting bee. Bee World 16, 111-113

Casteel DB (1912) The behavior of the honeybee in pollen collecting. US Bur Entomol BullNo 21

Chandran K, Suryanarayana MC (1970) A noteon entomophily in Dendrophthoe falcata, andits usefulness to the Indian honeybee. J ApicRes 9, 45-47

Chaturvedi M (1973) An analysis of honeybeepollen loads from Banthra, Lucknow, India.Grana 13, 139-144

Chaturvedi M (1977) Further investigation onthe pollen analysis of bees loads from Ban-thra, India. New Bot 4, 41-47

Chaudhari RK (1977) Forage in Punjab plains(India). Pathankot and adjacent villages. Indi-an Bee J 39, 15-20

Darade MN, Singh BP, Chakrawarty P (1989)Flight range of Apis cerana. Indian Bee J 51,59-61

Deodikar GB (1965) Melitto-palynology. IndianBee J 27, 59-72

Dietz A (1975) Nutrition of the adult honey bee.In: The Hive and the Honey Bee. Dadant andSons, Hamilton, IL

Erdtman G (1952) Pollen Morphology and PlantTaxonomy-Angiosperms. Almquist and Wik-sells, Uppsala

Hodges D (1955) Pollen loads, the way theworker bee gathers, packs and stores themfor use. Am Bee J 95, 475-482

Knee WJ, Moeller FE (1967) Comparative studyof pollen sources of honeybees and bumblebees. J Apic Res 6, 133-138

Louveaux J (1954) Études sur la récolte du pol-len par les abeilles. XV Int Beekeep Congr,Copenhagen

Louveaux J (1958) Recherches sur la récolte dupollen par les abeilles (Apis mellifica L). AnnAbeille 1, 113-188, 197-221

Maurizio A (1949) Pollenanalytische Untersu-

chungen an Honig und Pollenhöschen. BeihSchweiz Bienenztg 2, 320-455

Maurizio A (1953) Weitere Untersuchungen anPollenhöschen. Beih Schweiz Bienenztg 2,486-556

Maurizio A (1960) Bienenbotanik. In: Biene undBienenzucht (Büdel A, Herold E, eds) Ehren-wirth Verlag, Munich

Naim M, Phadke KG (1972) Flight range of Apiscerana indica when enticed to a feedingdish. J Apic Res 11, 167-169

O’Neal RJ, Waller GD (1984) On the pollen har-vest by the honey bee (Apis mellifera L) nearTucson, Arizona (1976-1981). Desert Plants6, 81-94, 99-109 (AA 558/1986)

Percival M (1947) Pollen collection by Apis mel-lifera. New Phytol 46, 142-173

Phadke KG, Naim M (1974) Observations onthe honeybee visitation to the litchi (Nepheli-um litchi) blossoms at Pusa (Bihar, India). In-dian Bee J 36, 9-12

Rao GM (1983) Litchi and bee-keeping. IndianHortic 28 (2), 19-20

Rao GM, Nair KS (1985) Mango tree - a sourceof bee forage. Indian Bee J 47, 49-50

Severson DW, Parry JE (1981) A chronology ofpollen collection by honeybees. J Apic Res20, 97-103

Sharma M (1970a) An analysis of pollen loadsof honey bees from Kangra, India. Grana 10,35-42

Sharma M (1970b) Studies on the pollen loadsof honeybees from Kangra, India. J Palynol6, 104-110

Sharma N (1983) Melissopalynology and surveyof honey plants in Simla Hills (Himachal Pra-desh). MPhil Project Report, Himachal Pra-desh University, India (AA 1245/85)

Singh TSMS, Joshi MA, Suryanarayana MC(1987) Evaluation of sources of pollen to hon-ey bees at Vijayarai (Andhra Pradesh). Proc5th All India Symp Palynol, Nagpur, India,65-71

Sladen FWL (1911) How pollen is collected

by the social bees and the part played inthe process by the auricle. Br Bee J 39, 491-506

Suryanarayana MC (1975) Studies on bee bota-ny and palynology of the flora and coorg andadjacent parts of the Mysore State. Ph D the-sis, Poona University

Suryanarayana MC (1978) Bee plants of India.1. Carvia callosa (Nees) Brem. Indian Bee J40, 7-10

Suryanarayana MC, Singh TSMS (1989) Parthe-nium hysterophorus - a pollen source to Apismellifera. Bee World 70, 127-129

Suryanarayana MC, Rao GM, Phadke RP

(1987) Higher yields of sunflower throughhoney bees. Indian Fmg 37 (2), 5-7, 19

Suryanarayana MC, Seethalakshmi TS, PhadkeRP (1981) Pollen analysis of Indian honeys -1. Honeys from litchi (Nephelium litchi)and Jamun (Syzygium cumini). Proc IV IntPalynol Conf, Lucknow 1976-1977, 3, 491-498

Synge AD (1947) Pollen collection by honeybees. J Anim Ecol 16, 122-138

Todd FE, Bretherick O (1942) The compositionof pollens. J Econ Entomol 35, 312-317

Wodehouse RP (1935) Pollen Grains. McGrawHill, New York